Evaluating the Subsistence Service Recovery: Spatial and Temporal Characterization of Prince William Sound Subsistence Harvest Activities

Final Report

Aaron Poe National Forest 3301 C Street Suite 300 Anchorage, 99503

Dr. Randy Gimblett BSE Bldg. Rm. 325 School of Natural Resources and Environment University of Arizona Tucson, Arizona 85712

Milo Burcham Cordova Ranger District P.O. Box 280 Cordova, Alaska 99574

December, 2010

“The Valdez Oil Spill Trustee Council administers all programs and activities free from discrimination based on race, color, national origin, age, sex, religion, marital status,pregnancy, parenthood, or disability. The Council administers all programs and activities in compliance with Title VI of the Civil Rights Act of 1964, Section 504 of the Rehabilitation Act of 1973, Title II of the Americans with Disabilities Act of 1990, the Age Discrimination Act of 1975, and Title IX of the Education Amendments of 1972. If you believe you have been discriminated against in any program, activity, or facility, or if you desire further information, please write to: EVOS Trustee Council, 441 West 5th Avenue, Suite 500, Anchorage, Alaska 99501-2340; or O.E.O. U.S. Department of the Interior, D.C. 20240

2 Evaluating the Subsistence Service Recovery: Spatial and Temporal Characterization of Prince William Sound Subsistence Harvest Activities

Final Report

Aaron Poe Chugach National Forest 3301 C Street Suite 300 Anchorage, Alaska 99503

Dr. Randy Gimblett BSE Bldg. Rm. 325 School of Natural Resources and Environment University of Arizona Tucson, Arizona 85712

Milo Burcham Cordova Ranger District Chugach National Forest P.O. Box 280 Cordova, Alaska 99574

December, 2010

3 Evaluating the Subsistence Service Recovery: Spatial and Temporal Characterization of Prince William Sound Subsistence Harvest Activities

Final Report

Study History: This characterization of harvest patterns was initiated in 2008 to provide land and resource managers with insights into the recovery of the human service of subsistence harvest in the aftermath of the Oil Spill (EVOS). It was initiated to define seasonal harvest distribution, relative use levels, and overlap with other human activities within Prince William Sound in order to evaluate how the recovering service of subsistence relates to other human activities within the region. It tiers to a suite of projects launched by the Chugach National Forest (CNF) to evaluate the dynamics of human use in the Sound as they relate to recovering injured resources and services.

Abstract: We created a mapped distribution characterizing seasonal harvest activities of subsistence eligible households in Prince William Sound. Our analysis focused on data collected during interviews with 88 harvesting households in the communities of Chenega Bay, Cordova, Tatitlek and Whittier in 2009. Harvest efforts were reported for 24 individual resources and were summarized in terms of: 1) proportion of households using individual resources; 2) spatial extent of harvest; and 3) relative intensity of harvest effort. We also evaluated total reported use relative to recent use and found slight declines in overall household harvest effort. A spatial characterization of use intensity was compared to predicted distributions of recreation use activities throughout PWS and a weak positive correlation was detected between these activities during summer months. The results of this assessment will allow land and resource managers to make more informed decisions relative to protecting subsistence eligible harvesters, thus promoting the recovery of the Subsistence Human Service in the region that was injured by the oil spill.

KeyWords: Prince William Sound, Subsistence, Household survey, Harvest, Alaska, Chenega Bay, Cordova, Tatitlek, Whittier

Project Data: Data were collected using household surveys and entered into a database for further analysis. Upon completion of the project, harvest data pooled from all communities (devoid of any individual information) and summarized into GIS layers will be housed at the Chugach National Forest, Supervisor’s Office, Anchorage Alaska.

Recommended Citation: Poe, A.J., H. R. Gimblett and M. Burcham. 2010. Evaluating the Subsistence Service Recovery: Spatial and Temporal Characterization of Prince William Sound Subsistence Harvest Activities, Restoration Project Final Report. USDA Forest Service, Chugach National Forest, Anchorage, Alaska

4 Table of Contents Page No. Figures and Tables 6 Executive Summary 8 Introduction 9 Relevance to 1994 Restoration Plan Goals and Scientific Priorities 11 Project Design 12 Objectives 12 Study Area 12 Methods 15 Household Surveys 15 Interview Methods 16 Analysis 16

Results 17 Household Surveys 17 Spatial Analysis 29

Discussion 36 Conclusions and Recommendations 39 Acknowledgments 40 Literature Cited 40 Appendices 43

5 List of Figures Page No. Figure 1 - Prince William Sound Study Area 13 Figure 2 - The vector of intended use change in number of 19 records (ie., polygons where used) of 24 resources by 88 households in Prince William Sound. Figure 3 - The vector of intended use change in number of 19 harvest records (ie., polygons where used) of four resources experiencing the greatest declines in household use in Prince William Sound. Figure 4 - Total harvest days invested annually in Prince William 21 Sound between ~2004-2008 by residents from Chenega Bay. Figure 5 - Total harvest days invested annually in Prince William 21 Sound between ~2004-2008 by residents from Cordova. Figure 6 - Total harvest days invested annually in Prince William 22 Sound between ~2004-2008 by residents from Tatitlek. Figure 7 - Total harvest days invested annually in Prince William 22 Sound between 2004-2008 by residents from Whittier. Figure 8 - Annual total of harvest days from Prince William Sound 25 Communities from January through December during the period from ~2004-2008. Figure 9 - Seasonal harvest day efforts for individual species 26 harvested by 88 households in Prince William Sound. Figure 10 - Fourteen sub-regions of Prince William Sound 28 identified by two or more households as areas they intend to use less as a result of competition with others. Figure 11 - Total number of reports (n=55) from respondents, 29 by species and community, who identified 37 unique sub-regions they now intend to harvest less due to competition with others in Prince William Sound. Figure 12 - Total annual harvest day density for sub-region 31 polygon units broken into four classes from lighter to darker representing low, medium, higher, with highest being highlighted in blue. Figure 13 - Total spring harvest day density for sub-region 32 polygon units broken into four classes from lighter to darker representing low, medium, higher, with highest being highlighted in blue. Figure 14 - Total summer harvest day density for sub-region 33 polygon units broken into four classes from lighter to darker representing low, medium, higher, with highest being highlighted in blue. Figure 15 - Total fall harvest day density for sub-region polygon 34 units broken into four classes from lighter to darker representing low, medium, higher, with highest being highlighted in blue.

6 Figure 16 - Total winter harvest day density for sub-region polygon 35 units broken into four classes from lighter to darker representing low, medium, higher, with highest being highlighted in blue.

List of Tables Page No: Table 1 - The proportion of 88 households reporting use of a 18 resource through time and within the past five years (~2004-2008) in Prince William Sound. Table 2 - Annual harvest days reported by respondents from 20 four Prince William Sound communities by resource type between ~2004-2008. Table 3 - Average years of experience harvesting resources held 23 by respondents from four Prince William Sound Communities. Table 4 - Average history of experience in years of harvesting 23 individual resources held by 88 households from Prince William Sound. Table 5 - The seasonal density of harvest days/km2 reported 30 by respondents from 88 households in Prince William Sound.

7 Evaluating the Subsistence Service Recovery: Spatial and Temporal Characterization of Prince William Sound Subsistence Harvest Activities

EXECUTIVE SUMMARY The Chugach National Forest, as the major land-owning Federal Trustee in the Sound, plays an important role in the EVOS recovery process. One area of critical importance to Forest managers, which has received less attention by researchers, is the distribution, behavior, and experience of human users throughout the Sound and the impact of these users on EVOS recovering resources and services. A key recovering human service is subsistence harvest of resources by eligible PWS communities which include Chenega Bay, Cordova, Tatitlek and Whittier. Recreation use is increasing in the Sound, and there is concern that increased competition from commercial and independent recreation may be negatively impacting harvest activities from these communities through direct competition for resources. This competition could be coming directly through taking of resources by sport fishers and hunters, and also indirectly through displacement of subsistence harvesters from traditional harvest areas.

We conducted interviews of 88 households in Chenega Bay (8), Cordova (35), Tatitlek (6) and Whittier (39) in 2009. We asked households to summarize their harvest activities within the Prince William Sound region in terms of resources sought, spatial extent of harvest and total numbers of days invested. We examined changes in harvest activity by asking respondents to summarize harvest based on a total history of use and report on their household’s activities during a recent five year period from ~2004-2008. We also asked them to summarize their intentions for harvest in future years by identifying areas they may use less or abandon completely. For those areas identified we asked respondents to identify motivations for likely decreases in use.

All information collected was summarized by season of use and spatially, across a series of 160 terrestrial and marine polygons. The resulting information was entered into a GIS for summary and analysis. We evaluated resulting spatial and seasonal distributions of harvest effort with predictions of recreation activity density from a concurrent study evaluating recreation user experience in the Sound.

We documented substantial investments in resource harvest by households from subsistence eligible communities in terms of days invested and spatial extent of travel in pursuit of resources. Harvesting of 24 different resources was reported, primarily halibut, salmon, rockfish, berries and deer. All harvestable resources identified have rich histories of use by respondents but all except salmon show apparent declines in use during the past five years. This decline was evidenced by a decrease in the number of households reporting use of the resource as well as a decrease in spatial extent of resource harvest. This decrease in reported use was paralleled by reported intentions to reduce use within certain sub regions within the study area.

Reasons for decreased use were primarily related to a variety of personal circumstances related to lifestyle change, aging/medical, household move, etc. but a close secondary reason stated was an apparent decline in resources within individual sub regions

8 (polygons). The third and fourth most often stated reasons were competition from other users followed by transportation costs.

Respondents reported competition within 37 individual polygons (~23%) that would likely result in them using those areas less. No respondents reported the intention of abandoning use of areas completely due to competition with others. When asked to rank who their perceived competition was, respondents overwhelmingly identified other harvesters as their main competition. They perceived these competitors as most likely to be other local harvesters, followed by commercial sport (guided hunting and fishing), and non-local sport, and commercial harvesters (commercial fishing). We found a weak positive correlation between higher densities of recreation activity and areas favored for subsistence harvest during summer. This indicates that the overall interaction with recreation may be greatest during summer but it’s not completely clear how this interaction may result in feelings of competition.

Perhaps most important is our effort to define a baseline for spatial distribution and seasonal intensity of subsistence harvest in the Sound. Future management efforts, investigations and monitoring will likely benefit as a result. Some key recommendations being made as a result of this work include:

INTRODUCTION In 1989, Prince William Sound (hereafter, “the Sound”), the heart of the Chugach National Forest (CNF), was severely impacted by the Exxon Valdez Oil Spill (EVOS). In the aftermath of the spill, federal and state trustees were awarded criminal and civil restitution funds to help with the recovery (and the evaluation of the recovery) of injured resources and human services, including traditional practices of subsistence harvest, which is still listed1 as “recovering” (EVOS Trustee Council 2010).

The original cause for concern relative to subsistence was that contamination from oil has altered the availability of safe subsistence resources in the Sound. For example in Chenega Bay, prior to the spill, marine mammals made up about 40% of the subsistence harvest and only 3% ten years later in 1998. Ten years later, 63% of the households in oil spill affected areas believe that subsistence resources have not recovered from the effects of the spill. Some subsistence community members reported having to increase effort (traveling farther, spending more time and money) to achieve comparable harvests to those before the spill and an increasing reliance on fish (Fall 1999).

Subsistence harvest is a critical part of the rural Alaskan lifestyle. It is widely recognized in Alaskan land and wildlife management that subsistence harvest provides irreplaceable cultural, spiritual, personal, and sustenance value. Alaska National Interest Lands Conservation Act (ANILCA) requires the federal land managers consider the effects of management on subsistence activities (USDA Forest Service 2002a). People in the Alaska Native communities of Chenega Bay and

1 A complete list of injured resources and services is on the EVOS Spill Trustee Council’s website at: http://www.evostc.state.ak.us/Publications/injuuredresources.cfm.

9 Tatitlek as well as citizens of Cordova and Whittier partake in a variety of subsistence harvest activities in the Sound and qualify for various forms of federal subsistence harvest. Fish and marine mammals comprise the majority of subsistence resources taken but there is also significant use of other species including Sitka black-tailed deer, black bear, mountain goats, waterfowl, seabirds, river otters and mink (Stratton et al. 1986, Fall et. al. 1999). Although many of these species use terrestrial habitats the majority of time spent in pursuing their harvest occurs in direct proximity to marine, inter-tidal and shoreline areas. Because the majority of access to both upland and near shore subsistence resources occurs via small watercraft, subsistence activities related to upland species may also impact resources in the inter-tidal areas that serve as access points to inland subsistence gathering. Historically, subsistence communities have expressed concern that activities such as timber harvest, road building, and recreation development could impact fish and wildlife populations or increase competition for subsistence resources. In the Sound, subsistence users are concerned about increased competition for wildlife and fish resources from increasing numbers of private, urban users and commercial operations (J. Fall, ADF&G, Division of Subsistence, pers. comm. with A. Poe, 2007). Recreational use is not evenly distributed in PWS. Certain areas are more desirable for a variety of reasons including: distance from access communities, presence of tidewater glaciers and post-glacial landscapes, availability of landing beaches, protected anchorages, sport fish streams, wildlife viewing opportunities, cabins and wild game concentrations. Independent use is not well understood for PWS though the majority of use happens during June, July, and August (Murphy et al. 2004). Recent work completed by Poe et al. (2010a) produced predictive GIS raster surfaces for recreation activity in the Sound during spring, summer and fall. Subsistence harvesters in PWS are also concerned about increased commercial recreation activity interfering with their subsistence practices. For example many of the traditional harvest areas of the people of Chenega Bay, as characterized by Statton et al. (1986), have become popular for recreation activities and it is reasonable to assume that this may result in increased contact and competition between user groups. Given that some recreationists using PWS may not understand the harvest traditions and rights of subsistence users, the potential for conflict is significant. Conflicts between user groups have significant implications for EVOS impacted resources and services. Conflicts can diminish quality of life/experience for both subsistence and recreation groups (each already harmed by the spill) and push harvest and recreation activities into previously unused areas, potentially negatively affecting the 25 impacted and recovering resources

Various subsistence harvest studies have been conducted by researchers from Alaska Departments of Fish and Game (ADF&G) Division of Subsistence since the early 1980s (e.g., Fall et al. 1996). The results of these surveys are used to update a community profile database maintained by the Division (Scott et al. 2001). Unfortunately the variables recorded in such surveys do not allow for spatially explicit analysis of harvest effort characterized in this database. The database tracks numbers of individuals (e.g., marine mammals) or pounds of resources (e.g., salmon) harvested but does not describe

10 the daily effort of harvest nor location of harvest at any level of resolution finer than all of PWS. This type of information is for a few hunted game species whose harvest is managed by ADF&G under the statewide harvest reporting effort though typically the communities of Tatitlek and Chenega Bay do not participate in this survey (D. Crowley, ADF&G Area Biologist for PWS, pers. comm. with A. Poe, 2010). Furthermore, perceptions about resource availability, conflict with other users or abandonment of traditional areas is more difficult to evaluate with these data sources.

One study completed in the early 1980’s following the resettlement of the village of Chenega Bay (the original village site was destroyed in Alaska’s 1964 Good Friday Earthquake) evaluated a change in distribution of harvest and did report some mapped results of harvest areas. The data focused only on households from Chenega Bay and results are considered to be out of date at this time based on reported changes in harvest location as a result of the oil spill (J. Fall, ADF&G, Division of Subsistence, personal communication with A. Poe 2007). Consultation with PWS native communities and regional subsistence managers about evaluating the distribution of harvest efforts has confirmed this information is currently lacking for two native communities of Chenega Bay (Chenega Corporation 2008) and Tatitlek (Tatitlek IRA Council 2003).

Though spatial characterization has been completed for recreational use in the Sound (e.g. Murphy et al. 2004, Gimblett and Itami 2006, Poe et al. 2010a) less is known about the spatial and temporal nature of subsistence harvest activities in PWS. A key resource offering insights into harvest activity is the An earlier study compiled spatially explicit harvest information for spring black bear sport harvest in the western Sound (Lace et al. 2008) but a comprehensive, spatially explicit effort for multiple species has not be completed. Understanding the subsistence harvest patterns in the Sound adds critical depth to the few existing PWS human use studies by describing the exact nature of overlap between subsistence and recreation use in Sound. This information will help managers better understand the dynamics around these potential interactions and anticipate potential conflicts between these two injured human services.

Relevance to 1994 Restoration Plan Goals and Scientific Priorities The EVOS Trustee Council maintains the status of subsistence activities as a recovering human service in the Sound. The primary cause for concern is that contamination from oil has altered the availability of safe subsistence resources in PWS. Given the existing injury to subsistence harvest and its associated resources following the spill, competition with sport hunters / fishers, and the potential displacement from favored harvest areas by increased recreation, there is considerable potential to produce a cumulative negative effect on the subsistence lifestyle of PWS communities as well as to recovering or recovered species. As the CNF and partner agencies manage recreation in the Sound, careful consideration of impacts to subsistence users should mitigate potential impacts to this important user group and the EVOS injured resources they depend upon.

It should be noted that this project has been proposed independently from a larger analysis aimed at evaluating the status of habitat restoration by restoring subsistence food safety confidence within communities throughout the area impacted by EVOS

11 (commonly discussed as the subsistence portion of “the re-opener”). However, our proposed objective and resulting data products have significant potential for integration with this larger effort. Though our focus is specifically on communities within PWS it may allow for the limited inference into the potential for harvest displacement resulting from interference by other human use activity. This potential constraint on subsistence harvest may be especially acute for PWS communities as a result of the dramatic increase of recreational use in the Sound in recent years. As such, our characterization of the overlap of subsistence and recreation activities may be an important variable that could improve this larger analysis as it pertains to PWS communities. Our products are spatial data layers summarizing harvest effort by season. The feature layers are based on harvest boundaries managed by the Alaska Department of Fish and Game (ADF&G) promoting maximum flexibility for inclusion into any future analyses.

Project Design This effort relies on first person household interviews conducted face-to-face within the communities of interest. These interviews depend on respondents being willing to share the seasonal harvest activities of their household via a generalized map document upon which total days of effort per resource is compiled. We will measure the amount of spatial and temporal overlap between private recreation characterized by activity type (kayak, sport harvester, pleasure boater) and subsistence harvesters. This analysis will produce a baseline of existing overlap between private recreationists and subsistence harvesters and will identify areas of potential conflict between the two activities. An additional assessment will be completed to evaluate existing overlap with commercial recreation and subsistence harvest based on Special Use Permit databases kept by the Glacier and Cordova Ranger Districts. The structure of this data may not lend itself to a rigorous quantitative measure (like that from the Human Use Model) but would provide a qualitative assessment for recreation managers in PWS.

Objectives 1. Characterize spatial and temporal patterns of subsistence harvest and evaluate how other human use and resource management actions may overlap with harvest activities. 2. Provide insights into prioritizing and protecting the rights of subsistence users by understanding potential conflicts with recreation users, while still providing access for these non-subsistence groups. Provide insights on long term monitoring of subsistence uses relative to management activities as a means for continuing to evaluate potential competition or conflict between user groups.

Study Area Prince William Sound (PWS) is located in south-central Alaska at 61o N, 148o W (Figure 1). The Chugach and Kenai Mountain ranges separate most of PWS from interior Alaska and two large islands, Montague and Hinchinbrook, shelter the hundreds of bays and islands that make up PWS from the (Murphy et al., 2004). The maritime climate of PWS is characterized by heavy annual precipitation, much of which falls in the form of snow during long winters. Summers are generally cool and wet Lowlands are dominated by old-growth Western Hemlock (Tsuga heterophylla) and Sitka Spruce

12 (Picea sitchensis) forests. Terrestrial vegetation begins within 1-2 m of the high-tide line. Blueberries and huckleberry (Vaccinium sp.), Rusty Menziesia (Menziesia ferruginea), Devilsclub (Oplopanax horridum) and salmonberries and thimbleberries (Rubus spp.) are common understory species found in forests and disturbed areas. The shoreline is comprised of tall, rock cliffs, gravel beaches, tidal flats, rocky outcrops and islands, estuaries and tidewater glaciers. PWS shorelines are exposed to large fluctuations in tide (+6 m to -1 m) and different levels of wave action (Bowyer et al. 1995).

The Sound is predominantly remote and wild. Road access is limited to the communities of Whittier and Valdez, and the transportation corridors consist solely of marine waterways. CNF manages most of upland PWS, including the 2.1 million acre Nellie Juan Wilderness Study Area (Twardock and Monz 2000). The state of Alaska as well as the Chugach Alaska, Chenega, Tatitlek and Eyak Alaska Native corporations own approximately 20% of land in PWS, with another ~1% being privately-owned.

Figure 1: Prince William Sound study area

According to Fall et al. (1996) in PWS the ”..subsistence cycle begins in the spring with harvests of herring and herring spawn as well as halibut, Dolly Varden, rockfish, smelt, and cod. Spring harvests also provide invertebrates (such as clams, octopus, and chitons), birds, eggs, harbor seals, and sea lions. Summer is traditionally the busiest time of year, when people harvest and preserve large quantities of salmon for winter use. Autumn also is important for salmon fishing, as well as for hunting, gathering marine invertebrates, and harvesting wild plants. Subsistence activities in winter include hunting, marine fishing, gathering chitons, clams, and other invertebrates along the beaches, and trapping.” The Division of Subsistence within ADF&G maintains records of systematic household surveys for PWS communities reporting subsistence harvest within the Community Profile Database (Scott et al. 2001).

13

Recreational use is not evenly distributed in PWS. Certain areas are more desirable for a variety of reasons including: distance from access communities, presence of tidewater glaciers and post-glacial landscapes, availability of landing beaches, protected anchorages, sport fish streams, wildlife viewing opportunities, cabins and wild game concentrations. Independent use is not well understood for PWS though the majority of use happens during June, July, and August (Murphy et al. 2004). The majority of use takes place within ½ mile of the saltwater shoreline, which is primarily accessed by kayak or motorboat. Upland use is limited, due to the boggy nature of most soils above the beaches and due to steep, rugged terrain. Activities are mostly low impact, such as small group overnight use, day hiking, sightseeing, hunting and fishing. Contemporary recreation use studies completed by Poe and Greenwood (2010) and Poe et al. (2010a) have defined relative intensities of recreation and other human use patterns in the region.

In the twenty years since the Exxon Valdez oil spill (EVOS), Prince William Sound has experienced numerous changes. The spill itself impacted and disrupted the resources and human services in the Sound. For example, it is believed that recreation/tourism services were redistributed from oil impacted areas of the Sound to areas of the Sound that had previously experienced little use (Hennig and Menefee 1995). Over the past decade, the Sound has also experienced increased human use. The growth of the recreation/ tourism sector statewide has been accompanied by improved access to the Sound (Colt et al. 2002). In the western Sound in particular, the opening of the Whittier access road in 2001 has led to both increased personal and commercial recreation/tourism use (Fay 2008).

Recreation use of the Sound is strongly seasonal due to variable weather and extreme winter conditions. Recreational use of uplands (including sport hunting), recreational boating, and sport fishing occur mainly from May through September. Previous studies have used a recreational season of May 1 – September 30 in eastern PWS (Wolf et al. 2006). Four seasons were defined for the Prince William Sound Human Use Hot Spots GIS Database and Spatial Analysis project (Poe and Greenwood 2010). They are:  Spring = April 1 thru June 14  Summer = June 15 thru August 31  Fall = September 1 thru December 31  Winter = January 1 thru March 31

These seasons were defined based on a review of existing recreational use studies of PWS as well as sport hunting seasons established by ADF&G for key species harvested under sports regulations including brown bear, black bear, deer, and mountain goats (Poe and Greenwood 2010).

Residents from the four rural Prince William Sound communities (Chenega Bay, Cordova, Tatitlek, and Whittier) that qualify for federal subsistence activities, as provided by the Alaska National Interest Lands Conservation Act (ANILCA) are the focus of this effort. These household style interviews about the subsistence harvest patterns have proven to be successful at evaluating summary harvest of resources in the years prior to and following the 1989 Exxon Valdez Oil Spill (e.g., Fall et al. 1996). This population is

14 comprised of approximately 1107 households: Cordova (958), Whittier (86), Tatitlek (38), and Chenega bay (22) according to 2000 Census records. Approximately 20% of those individuals are from three different Alaska Native communities, commonly referred to as Chenega, Tatitlek, and Eyak (whose residents are dispersed throughout the community of Cordova).

METHODS A two step process was implemented in order to evaluate harvest effort overlap with recreation use. The first involved first-person, face-to-face household interviews and the second was translating those results into spatially and seasonally explicit characterizations of harvest effort. It should be noted that the following general ethical principles (established as the appropriate way to engage subsistence communities in ) guided our efforts: “1) review and approval of the research plans by community governments prior to fieldwork; 2) informed consent by household members selected for interviewing (participation in the research was voluntary), 3) confidentiality of individual and household-level responses, 4) review of study findings by the participating communities and 5) providing study findings and reports to each study community” (Fall et al. 1999). These principles are consistent with those developed by the EVOS Trustee Council in regard to “Protocols for including Indigenous Knowledge in the Exxon Valdez Oil Spill Restoration Process.”

Household Surveys Household style interviews around the theme of subsistence harvest have proven to be successful at evaluating summary harvest of resources in the years prior to and following the 1989 Exxon Valdez Oil Spill (Fall et al. 1996; Fall et al. 1999). However, as a prudent first step five members from three of the target communities (Cordova, Tatitlek and Whittier) were engaged in a pretest of the interview question and data summary tools. The intent of this pretest was to identify and clarify any potentially confusing questions, evaluate appropriateness of categorical responses, and solicit general comments regarding the proposed interview approach and the appropriateness of the associated map document used to record responses.

Following survey instrument improvement, residents from the four communities of PWS who are subsistence eligible (Chenega Bay, Cordova, Tatitlek, and Whittier) were consulted through individual household interviews conducted by current community institutions (e.g., tribal or community councils) or trusted residents from those communities. In the case of the two native villages of Tatitlek and Chenega individuals were approved by tribal leadership. Interviews in Cordova were conducted by employees of the Native Village of Eyak. The remote nature of these communities made this approach the most effective for reaching all households.

Information was collected through oral interviews and results were recorded by interviewers with assistance from a map document. Interviews occurred between March and November of 2009 in all four communities. Respondents described their household’s harvest practices for each resource in terms of the number of years pursing each resource;

15 the month(s) of harvest and the duration of days invested in each resource. They were asked to summarize these efforts relative to spatial polygons representing sub-regions of Prince William Sound. For land-based harvests, they reported harvest using 88 Uniform Code Unit (UCU) boundaries designated by ADF&G for the uplands of the PWS region. Water based harvest was summarized to 72 marine polygons used to summarize sport fish harvest by the ADFG Sport Fish Division. Respondents were also asked about any recent changes in the location and effort expended in their harvest activities due to other users or changes in resource availability.

Responses were summarized and analyzed primarily as a summary dataset for PWS as a whole. The spatial and temporal distribution of harvest was summarized by seasons (spring 4/1 – 6/14; summer 6/15 – 8/31; fall 9/1 – 12/31; winter 1/1 – 3/31) and compared to predicted distributions of recreation use by private individuals during those same seasons.

Interview Methods Phone records from the municipality of Cordova were used to randomly select 340 (~40% of total) households for interviews based on recommended sampling procedures for Alaskan communities of < 2000 households (State of Alaska 2008). In the smaller communities of Chenega Bay (21 households), Tatitlek (38), and Whittier (86) (2000 Census data) a combination of phone lists and door-to-door visits were used to conduct a complete census of 146 households (Fall et al. 1999). Interviewers introduced the project and made an inquiry by phone or in person about the availability of the head of each household to meet for a face-to-face interview for approximately 30 minutes to answer questions about their subsistence harvest patterns. Three separate attempts were made to contact each household. Interviewers arranged meeting locations amenable to the head of household with an emphasis on the convenience of the respondent.

A total of five categorical response questions, including one four-part question, and two narrative response questions were asked during the course of the interview (See Appendix I). Interviewers concluded with reviewing brief review of all reported results with the respondent to capture any inconsistencies in response or transcription errors.

Analysis Harvest effort by season and species will be summarized relative to PWS as a whole. Summaries will include three primary indicators. The first measure is overall household use which is defined as any use of an individual resource reported by a surveyed house hold. The second measure is the spatial extent of harvest by household, defined as a water or land sub-region polygon where an individual resource was harvested. The third measure, representing intensity of effort, is harvest days (one harvest day being defined as any amount of time in a 24-hr period spent attempting to harvest an individual resource within any portion of a given sub-region). The variance around these three indicators of harvest effort, were summarized by season, resource, and community using descriptive statistics.

An analysis of household use over time was completed by comparing total history of use with recent use (in the last five years). Additional summary analyses of variance in total

16 harvest days invested by individual respondents will be completed for each harvested resource. The total of all harvest days was compiled by resource and season for each polygon and was divided by the total area (km2) of each polygon to return an index of harvest density. The 71 marine polygons were used as a zonal dataset in Spatial Analyst, Zonal Statistics, (ESRI 2009) to sample density raster surfaces predicting water based recreation activity in PWS by season (Poe et al. 2010a). An evaluation of overall correlation between recreation density and harvest density was completed using a Pearson’s Coefficient of Correlation analysis in Microsoft Excel 2007. Additionally, a rank summary analysis of sum recreation density relative to summary days of harvest effort will return those polygon units with the most intensely harvested polygons The results of this analysis predicts zones of highest intersection with recreation activity by season as well as the harvesting activities which most commonly co-occur with recreation activity.

A ranked summary analysis of most commonly selected reasons for change in use and discontinuation in use of cells was completed. A spatial summary of sub-regions (polygons) where respondents report changing their use was completed with results summarized by cited reason for change in activity. This same spatial, ranked-summary descriptive analysis was also completed for those cells where respondents reported discontinuation of use. Sub-regions in the Sound where conflict was reported by respondents where then compared to those sub-regions found to have high levels of intersection for both recreation and reported household harvest.

RESULTS Household Surveys A total of 88 households contributed 1473 records indicating individual locations of subsistence harvest reported by each household. Approximately 75% of those records were locations that had been used at least once during the period of ~2004-2008. The proportion of households within the community that contributed data ranged from 45% for Whittier, to 38% and 16% for Chenega Bay and Tatitlek respectively.

Cordova was approached through a sample selection scheme and attempts were made to contact ~90% of the 340 households originally targeted by the effort. Of this total population of 307 households, interviewers made contact (either spoke in person or exchanged voicemail messages) with 62% of household heads during up to three contact attempts. Of those 190 households that were successfully contacted, 25% refused to give any information regarding their household’s harvest efforts. Another 44% stated that they did not harvest resources within the Prince William Sound region. A further 14% reported they were harvesters but refused the survey. Of the remaining households contacted 34 completed surveys. This represents 18% of the original sampled population approached or 60% of those households who responded that they do harvest in the Sound.

All 88 households responding to the survey reported recent use within the period of ~2004-2008. When evaluating frequency of resource use by households, the resources most regularly used both throughout time and in recent years were: halibut, salmon, berries, rockfish and deer. Overall reported harvest appears to have decreased for a number of resources when comparing recent to total history of household use. The

17 average decrease in use for the 24 resources is 2.65% (+/- 2.25%) (Table 1). Decreases in household use of halibut, rockfish, deer, and shrimp are outside of the standard deviation we might expect in overall change in use and thus may represent actual declines in use of those resources. Households reporting harvest in recent years (~2004-2008) were asked about intended future use of individual sub-regions (polygons). Of this group 54% of respondents reported the intent for the same amount of use, 37% report decreased and 8% report increasing use. Figure 2 further explores the change in resource use in terms of spatial extent of harvest as indicated by change in number of polygons of reported use. Figure 3 is a similar summary for those four resources which appear to be experiencing the greatest declines in household use.

Table 1: The proportion of 88 households reporting use of a resource through time and within the past five years (~2004-2008) in Prince William Sound. # of # of Households Households Using in last % Change Resource Using % 5 years % in 5 years Berries 56 64% 52 59% -5% Black Bear 8 9% 7 8% -1% Brown Bear 4 5% 3 3% -1% Crabs 12 14% 7 8% -6% Deer 44 50% 39 44% -6% Eggs 8 9% 5 6% -3% Furbearers 1 1% 1 1% 0% Halibut 66 75% 60 68% -7% Harbor Seal 9 10% 8 9% -1% herring spawn 1 1% 1 1% 0% Mountain Goats 10 11% 8 9% -2% Other 6 7% 6 7% 0% Other Finfish 21 24% 18 20% -3% Plants 19 22% 18 20% -1% Rockfish 50 57% 45 51% -6% Salmon 65 74% 62 70% -3% Sea Lion 2 2% 2 2% 0% 4 5% 4 5% 0% Seabird 1 1% 1 1% 0% Seaducks 9 10% 7 8% -2% Shellfish 27 31% 23 26% -5% Shorebirds 2 2% 1 1% -1% Shrimp 37 42% 32 36% -6% Waterfowl 21 24% 17 19% -5%

18

Figure 2: The vector of intended use change in number of records (ie., polygons where used) of 24 resources by 88 households in Prince William Sound.

Figure 3. The vector of intended use change in number of harvest records (ie., polygons where used) of four resources experiencing the greatest declines in household use in Prince William Sound.

When asked the reasons for decrease and abandonment of an area the majority of respondents identified other which was a mix of reasons ranging from health and aging to changes in lifestyle and a number concerned about lingering oil from the spill. The second and third most stated reasons for decrease and abandonment were general change in resource availability, number of reasons broadly summarized as competition with others, and cost of transportation.

19 When considering only those harvest efforts occurring within the past five years, 88 households reported investing 16723 total harvest days throughout the year across 126 land and water units (~80% of the sub-region polygons) to harvest one to several of 24 different resources (Table 2). Fish species including salmon, halibut and rockfish represent ~ 55% of the effort invested in harvest within the Sound. Another ~23% of total effort centers around resources acquired in the marine or intertidal environments likely meaning that over ¾ of total effort spent on harvest in the region occurs on or near the salt water. Berry picking and deer harvest are the most significant activities happening in the upland, making up 17% of the PWS harvest effort. The reported harvest days were summarized for each species into 24 GIS shape files and can be made available for community evaluation, research and other management endeavors.

Table 2: Annual harvest days reported by respondents from four Prince William Sound communities by resource type between ~2004-2008.

Sum of Resource Harvest Days % Salmon 3767 27% Halibut 2466 15% Rockfish 2116 13% Berries 1629 10% Harbor Seal 1258 8% Deer 1181 7% Shrimp 1057 6% WaterFowl 607 4% Shellfish 440 3% Seaducks 357 2% Plants 336 2% Crabs 232 1% Black Bear 185 1% Other Finfish 177 1% Other 174 1% Sea Lion 162 1% Eggs 143 1% Seabird 126 1% Sea Otter 102 1% Furbearers 64 <1% Mountain Goats 52 <1% Brown Bear 50 <1% herring spawn 21 <1% Shorebirds 21 <1%

20

Chenega Bay respondents (n=8) reported the most total harvest days (8006) of any of the four Sound communities and also harvested the greatest diversity of individual resources (22). This was followed by Cordova (n=35) with 4125 harvest days invested in 17 different resources. Whittier respondents (n=39) reported a total of 3736 days attempting to harvest 12 resources with Tatitlek (n=6) reporting the smallest total effort of 856 harvest days spent across 13 resources. When controlling for total numbers of respondents between communities, Chenega Bay has the most individual effort, followed by Tatitlek, then Cordova and Whittier. Figures 4-7 describes harvest days by resource for each community.

Figure 4. Total harvest days invested annually in Prince William Sound between ~2004- 2008 by residents from Chenega Bay.

Figure 5: Total harvest days invested annually in Prince William Sound between ~2004- 2008 by residents from Cordova.

21

Figure 6: Total harvest days invested annually in Prince William Sound between ~2004- 2008 by residents from Tatitlek.

Figure 7: Total harvest days invested annually in Prince William Sound between 2004- 2008 by residents from Whittier.

Of the 88 households reporting harvest of 24 different resources the average number of years of experience harvesting individual resources was 15.5 (+/- 12.1) years. Tatitlek’s respondents appeared to have the longest tradition of resource harvest when compared other Sound communities though certainly the 1964 earthquake which destroyed the original village of Chenega plays an effect on harvest tradition for members of Chenega

22 Bay. Table 3 describes the average annual years of harvest by respondents from each community.

Table 3: Average years of experience harvesting resources held by respondents from four Prince William Sound Communities.

Average History Community of Use (years) Chenega 12.4 (+/- 8.9) Cordova 15.8 (+/- 13.6) Tatitlek 41.3(+/- 11.6) Whittier 16.2 (+/- 9.6 Grand Total 15.5 (+/- 12.1)

Table 4: Average history of experience in years of harvesting individual resources held by 88 households from Prince William Sound. Average history Standard Resource of use (years) Deviation Brown Bear 26.4 13.9 Harbor Seal 25.4 14.6 Deer 18.9 15.2 Salmon 17.9 12.6 Berries 16.3 11.5 Eggs 15.0 12.2 Halibut 14.7 10.8 Other Finfish 14.5 9.8 Other 13.9 9.0 Rockfish 13.6 11.2 Waterfowl 13.6 11.9 Plants 13.6 10.3 Shrimp 13.4 10.8 Shellfish 13.1 9.1 Black Bear 13.0 10.4 Seaducks 11.8 11.4 Crabs 11.7 8.6 Mountain Goats 11.4 15.8 Sea Otter 10.4 3.2 Sea Lion 9.5 0.5 Furbearers 9.5 10.6 Shorebirds 7.6 8.6 Grand Total 15.5 12.1

23 When history of use was evaluated relative to specific resources being used by more than one house hold in our respondent group we found the harvest of brown bear and harbor seal had somewhat longer traditions of use. As with deviation in overall harvest history by community, there was substantial deviation around individual resources. Table 4 describes history of harvest experience by resource type.

Harvesting activity happened throughout all months of the year in the Sound with respondents from all communities except Whittier reporting some level of harvest days for each month. When evaluated by individual month harvest appears to peak in July, August and September for all Prince William Sound respondents. The community of Chenega Bay appears to have the most regular seasonal harvesting presence in the Sound with no single month having less than 100 harvest days. Figure 8 shows the month distribution of annual harvest effort for each community beginning with January and continuing through December. Evaluation of seasonal harvest patterns show some resources are consistently harvested throughout all seasons of the year including halibut, rockfish, salmon and shrimp. Figure 9 describes total harvest effort for each individual species during four seasons.

24

Figure 8: Annual total of harvest days from Prince William Sound Communities from January through December during the period from ~2004-2008.

25

Figure 9: Seasonal harvest day efforts for individual species harvested by 88 households in Prince William Sound.

26 Not a single respondent reporting recent harvest (~2004-2008) stated that they no longer intended to use areas as a result of competition with others. However 15 households in this same group, (17%) reported 55 locations where they intended to harvest less due in part to, or entirely because of, competition with others. These 55 reports came from 37 unique sub-regions of the Sound. These 15 households were from all four surveyed Sound communities. Of these sub-regions, four were identified by three different respondents as places they will now use less with an additional 10 being identified by two respondents. Of this total for 37 unique sub-regions, 17 were upland units and the remaining 20 were marine areas. A total of 14 sub-regions were identified by two or more respondents as locations they intended to use less in future years (Figure 10). Eight of those sub-regions identified were marine and six were upland polygons.

Respondents from this group further specified the 11 resources which they intended to harvest less in these sub-regions due to competition. Of those deer and halibut were most often identified by all four communities as resources they intended to use less as a result of competition with other users. Figure 11 describes the species that respondents from each community plan to harvest less in the future due to competition with others.

The greatest amount of competition resulting in likely displacement was reported during summer and fall seasons. Respondents reported competition likely to result in less intended use during all seasons of the year from Chenega Bay. The remaining three communities only reported competition that resulted in less effort spent in areas during spring summer and fall. Cordova and Chenega Bay reported the greatest amount of competition occurring during summer while Tatitlek and Whittier reported summer.

When recent harvesters were asked to identify all sources of competition and to rank those individuals most likely to be competing for resources, respondents reported the following from most to least contributing to their sense of competition: local harvesters; commercial sport harvesters; non-local, private, sport harvesters; commercial harvest; and non-harvesters.

27

Figure 10: Fourteen sub-regions of Prince William Sound identified by two or more households as areas they intend to use less as a result of competition with others.

28

Figure 11: Total number of reports (n=55) from respondents, by species and community, who identified 37 unique sub-regions they now intend to harvest less due to competition with others in Prince William Sound.

Spatial Analysis Harvest was reported from ~80% of sub-region polygon units across the Sound. When harvest day effort for an entire year combined was evaluated, relative to size of individual polygons, the four most used units were adjacent to the towns of Chenega Bay, Whittier and Cordova. Overall harvest day density values ranged from 0.00-59.6 harvest days per km2 for a water unit immediately adjacent to the community of Chenega Bay, comprised of Sawmill Bay and the middle portion of Elrington Passage. This unit is consistently the most frequently harvested and represents a substantial outlier when evaluating ranges of overall density of use. When removed from the dataset harvest density effort averages from 0.07 – 0.55 harvest days/km2 . Overall summer season is the most intensely harvested followed by Fall and Spring; though variability in effort is certainly high between all sub-region polygons during all three seasons. Even with this high variation winter sees much less harvest overall.

Figure 12 represents total annual harvest density as classified into four classes using the natural breaks function in ArcMap. It returns four classes that can be conceptualized as: low, medium, higher, and highest in terms of harvest days/km2 . A number of areas adjacent to communities are used at the highest levels including Sawmill Bay/ mid- Elrington Passage unit as well as Evans Island which supports the community of Chenega Bay, two units adjacent to Cordova and Passage Canal (Whittier). Areas near Main and Eshamy Bays, west of Knight Island, also show up as higher use along with Bainbridge Pass, and Icy and Whale Bays.

29

Table 5: The seasonal density of harvest days/km2 reported by respondents from 88 households in Prince William Sound. Harvest days/km2 Season Min Max Average SD Spring 0.00 3.03 0.33 0.65 Summer 0.00 7.64 0.55 1.09 Fall 0.00 4.8 0.44 0.77 Winter 0.00 0.99 0.07 0.17 Annual 0.00 13.58 1.37 2.36

Spring harvest (Figure 13) has similar top four units as annual and summer but has newly identified units of higher use associated with Knight Island Passage as well as Derrickson, McClure, and King’s Bays. During summer months (Figure 14) a similar pattern of overall use with three of the four top units being the same but Main Bay (a key salmon fishing location) replaces a unit south of Cordova. Summer harvest is also generally more uniformly spread across numerous near-shore areas of the Sound in contrast to the patchier pattern displayed by total annual and spring classifications of harvest effort. Fall harvest (Figure 15) has similar overall patterns to summer but increased effort is demonstrated in areas associated with waters and uplands around Cordova causing Passage Canal to drop from the highest use class. A number of upland areas of importance appear in the medium density effort class across Hawkins, Latouche, Elrington, Bligh Islands as well as a pocket on the northern tip of Montague Island. Winter use is relatively low over all, only occurring in medium levels around Tatitlek and Cordova and low levels around Whittier. There is however significant winter harvest activity in the vicinity of Chenega Bay, with the region’s top four harvest units (Sawmill bay/Mid Elrington Passage, Bainbridge Passage, Icy & Whale Bays) all occurring in close proximity to the village.

30

Figure 12: Total annual harvest day density for sub-region polygon units broken into four classes from lighter to darker representing low, medium, higher, with highest being highlighted in blue.

31

Figure 13: Total spring harvest day density for sub-region polygon units broken into four classes from lighter to darker representing low, medium, higher, with highest being highlighted in blue.

32

Figure 14. Total summer harvest day density for sub-region polygon units broken into four classes from lighter to darker representing low, medium, higher, with highest being highlighted in blue.

33

Figure 15: Total fall harvest day density for sub-region polygon units broken into four classes from lighter to darker representing low, medium, higher, with highest being highlighted in blue.

34

Figure 16: Total winter harvest day density for sub-region polygon units broken into four classes from lighter to darker representing low, medium, higher, with highest being highlighted in blue.

An overall correlation between harvest day density and recreation density (as compiled by Poe et al. 2010a) shows a weak positive association between the two activities when summarized at the zonal level of sub-region polygons (r= 0.20). When evaluated seasonally the positive correlation between harvest and recreation activity becomes strongest during summer (r = 0.35) and substantially weaker during spring (r = 0.09) and fall (r = 0.05) respectively. Winter evaluations were not possible as no data exists that can be used to predict distributions of winter recreation; which has been generally characterized as limited in spatial extent and intensity (Poe and Greenwood 2010).

Specific water polygons where recreation intensity is relatively high (within the top 20% density range during summer) only overlap one area (Sawmill bay) where harvest density is at its highest during summer. Other overlap between high density recreation areas and harvest areas of high importance occurs in Passage Canal, Main Bay, and Eshamy Bay. A large number of medium importance and high recreation use occur in the Sound with most focused around Port Wells, Cochrane Bay, Culross Passage/Island and Port Nellie Juan. Small pockets of this type of interaction with medium units also occur near Cordova and on the southwest part of Knight Island.

35 An evaluation of areas reporting high and highest levels of commercial sport fishing activity (Poe and Greenwood 2010) showed overlap with one of medium importance to respondents on the northeast end of Hinchinbrook Island (Hawkins Island Passage). A polygon of high relative commercial sport fishing activity overlaps much of the area north of Hinchinbrook and extends out into Montague Island Passage. A similar polygon of highest commercial sport fishing encompasses much of southern ends of Prince of Wales, Elrington and Latouche Passages; south of Chenega Bay. This polygon abuts the highest use polygon for respondents during the summer, in Sawmill Bay/mid- Elrington Passage.

DISCUSSION Harvest of a variety of resources from terrestrial and marine environments is a key activity for residents of Prince William Sound. Subsistence harvest activities for Cordova, Chenega Bay, and Tatitlek have been well studied by ADF&G division of subsistence in the aftermath of the oil spill but this is an important activity for a number of households in Whittier as well. The total numbers of days invested by households as well as the distances travelled by many to harvest resources clearly underscore the importance of this activity to the communities of the Sound. It is important to note that the 8 households from Chenega Bay reported more total use in terms or harvest days invested than the other communities though Cordova reported the most total records of harvest. When controlling for numbers of respondents per community Chenega led substantially in terms of harvest days followed by Tatitlek, Cordova and Whittier. In a similar evaluation of extent of effort per household (ie., number of records) Chenega also led followed by Cordova, then Whitter and Tatitlek. However, Tatitlek reported the longest overall history of harvesting in Prince William Sound. This would have likely been the case for Chenega Bay as well had Chenega residents not been so greatly impacted by the 1964 earthquake and the subsequent tragic loss of so many residents.

Overall household use of harvested resources in Prince William Sound appears likely to decline in terms of the numbers of resources harvested but also in terms of the spatial extent of harvest. Of the 24 resources harvested, households reported that they intended to harvest only six at similar levels in the future. Assuming respondents follow through with these intentions, the remaining 18 resources can expect less harvest from 1-7% of the households in our survey. Four resources (halibut, rockfish, shrimp and deer) were identified by the most number of households (six from our sampled group) as resources they intended to harvest less in the future. Though this is admittedly a relatively small number of households it is worth noting that these four resources have been strongly used in the past by harvesters in Prince William Sound even following the oil spill according to community harvest records maintained by Scott et al (2001).

When use decrease by all households is evaluated in terms of spatial extent of harvest (ie., areas used by individual households) 37% of records were attributed the future use value of an intended decrease by respondents. Additionally, 12% of records reported that respondents intend to abandon use at areas all together. Perceived competition with others does not appear to be the strongest motivator for declines in use or abandonment. The most often identified impetus was a number of personal reasons (other) ranging from

36 lifestyle changes, household moves and health/aging. This reasoning was followed in rank by perceived decreases in general resource availability and thirdly by perceived competition from other users. This relative ranking of reasoning holds true for all households in general and when considering just those reporting recent use (between ~2004 and 2008).

No households reporting recent use stated intent to abandon any areas completely because of competition with others. However 17% of these households did report that they intended to use 37 individual sub-regions less in the future due to the presence of others. The majority of these sub-regions are adjacent to the communities of Cordova, Whittier, Tatitlek and Chenega likely due in part to the fact that these are some of the most heavily used areas by harvesters. In these instances locations of water based competition centered around halibut and rockfish while land-based competition focused on deer and berries.

Two additional areas were identified including one which is in the highest category for commercial sport fish use as identified by Poe and Greenwood (2010) in lower Port Bainbridge and the second at the south end of the Narrows. Given the intense pressure from commercial sport fish operators in the vicinity of Chenega Bay and Tatitlek competition in these areas is unsurprising; however, the resources specified near Chenega were crabs and sea lion. Those identified near Tatitlek included halibut but also harbor seal. Some surprising areas of reported competition likely to result in less use include the Naked Island complex (for deer), an area of southwest Knight Island and the lower end of Montague Straight between Latouche, Green and western Montague Island. Lower than expected populations of deer in the Naked Island Complex (pers. comm. D. Crowley, ADF&G Area Biologist) may be contributing to an increase sense of competition. In the case of these latter two sub-regions resources involved included sea lions and crabs.

With respect to stated intent for decreased use it is worth noting that 5 households offered 49 records indicating areas they intended to use less as a result of continued concerns about resource contamination from the oil spill. Additionally, a single house hold identified four locations they intended to abandon all together based on contamination from the spill.

Certainly intent to use less or abandon does not necessarily equate with non use in the future (Hall and Cole 2007) but these stated intentions parallel reported declines in overall resource use when comparing total use history to recent use from 2005-2009. The most common reported singular reason (outside of other) for intended decreases in use was a perception of general resource decline. Recent fluctuations in populations, bad weather and high gasoline prices may also some further investigation around those perceptions would prove useful (Poe et al. 2010).

There does appear to be a positive correlation between those areas being more intensely harvested and areas of higher density recreation and this varies by season. The strongest positive correlation between reported harvest activity and intensity of recreation use came during summer months (r = 0.35). We know from contemporary evaluations (eg., Poe et

37 al. 2010a, Poe and Greenwood 2010) that there is substantially more total use in Prince William Sound during summer months (mid June – August) so this could potentially explain the result. A confounding effect associated with this analysis is that the recreation density surfaces used cannot separate out harvesting vs. non-harvesting recreation. They were created using reported use locations of a variety of recreation groups including a significant number of fishers and also locations of any vessel not readily identifiable as supporting commercial fishing or larger tour vessels. As a result the uses being compared may in part be one in the same and thus suffer from an autocorrelation effect.

Further these recreation density surfaces can only portray relative densities of use disconnected from any behaviors of groups involved. Many studies of use displacement in wildland settings (CITATIONS) highlight the behaviors of offending groups being a significant variable in predicting displacement (Hall and Cole 2007). It is worth noting that of the 15 households reporting they intended to use areas less as a result of competition their perception of their primary competitors was that they were other local hunters, followed by commercial sport harvesters. Thus recreation activities that are obviously not harvesting likely have less effect in terms of potentially displaying harvesters from the region; suggesting that behavior of groups in these scenarios is a key component related to displacement potential.

This analysis of recreation activity interacting with household harvest may also be somewhat limited by our inability to look at recreation activity in terms of terrestrial sub- regions. However the majority of harvest reported and recreation use described by Poe et al. (2010a) in the Sound does appear to be marine based. This is certainly the case during summer months in terms of resource harvest (with the exception of berries) and recreation activities where greater than 95% of groups encountered occur on the water (Poe et al. 2010a).

An important note in regard to Cordova was that of the 190 households that were successfully contacted a greater proportion stated that they didn’t harvest resources from the Prince William Sound Region than those responding in the affirmative. Our interviewers in this community often heard that the more significant source of resources harvested by households came from the Copper River Delta region. Many respondents felt that investigating household use from that region would be of equal if not greater importance. The annual harvest survey efforts led by ADF&G would certainly contribute the numbers in terms of households and harvest effort for the majority of resources harvested on the Delta though the experiential questions associated with this study may be illuminating.

Certainly our efforts have significant limitations in terms of the degree of participation across communities. It would be difficult to argue that these results are statistically representative of any community, particularly in the case of Tatitlek where overall participation was limited compared to the other three communities. As a result our broad summaries of Sound-wide resource harvest may be more valid than those reported for individual communities. However it should be noted that our levels of participation

38 parallel those of other studies in the region, as harvesting households are often somewhat reticent to share this information with researchers and government agencies. This effort should be seen as offering a baseline of household use, spatial extent and relative intensity of harvest effort across four seasons. The fact that this information was compiled spatially should allow for ease of integration with harvest data tracked by ADF&G, future analyses of human use, proposed management actions, and help to structure future monitoring of household resource harvest in Prince William Sound.

Conclusions and Recommendations Though this study’s results have limitations in their ability to represent harvest efforts from individual communities and around singular resources we do offer a series of recommendations to land and resource managers in the Prince William Sound region.

 A primary reason stated for potential decrease of use in the future was a perception of general resource decline in the region. Managers should attempt to better understand perceptions held by long-term harvesters about resource decline and correlate them to existing harvest (for example those maintained by ADF&G) or other trend data. Such an effort could capitalize on what may be a very useful Traditional Ecological Knowledge (TEK) data source and an opportunity to more strongly engage citizens in resource management.

 The Sound supports significant subsistence and private, sport harvest activities as well as a thriving commercial industry for fish and game. Clearly some level of displacement is already occurring for local harvesters (those from Sound Communities) but affects on other harvesters are not known. Similarly the salient aspects around competition between harvesters, such as motivations, expectations, and general behavioral norms for each group are not known. A comprehensive effort should be made to understand dynamics around competition between these groups of harvesters that incorporates available data on sport and commercial harvest with sociological investigations targeting these three (or more) different groups in the Sound. Ideally such a study would be conducted in concert between land managers and ADF&G, responsible for management of fish and wildlife populations in the region. It should also be implemented by a third party who can be seen as unbiased in its investigation.

 Competition likely to result in displacement seems to be focused around other harvesting user communities. If managers are considering limitations on commercially guided activities as an approach to managing quality of user experience, acute evaluation should be made of those that harvest resources from the region.

 Additional effort should be invested in the community of Tatitlek to ensure complete coverage of harvest efforts from this community. A number of households were apparently missed during surveys and upon seeing results from household surveys from around the Sound the Tribal Council and residents may be interested in supporting addition characterization efforts for these households.

39

 Land and resource managers in the region should work to establish a dialogue with households from the four communities who participated in this study. These individuals have long-term knowledge of the resources in the region in general as well as significant investments in their sustainable use into the future. Efforts to engage with those households should include informal, periodic consultation with those households demonstrating the greatest intensity and extent of harvest. This should be paired with systematic efforts to repeat the efforts described herein in terms of spatial summaries of harvest effort and perceptions about resource availability and competition from others. Ideally such monitoring would be entered into as a partnership with each of the four communities involved.

Acknowledgments EVOS Federal Trustees, data collectors in communities; Coleen Brown; reviewers; Tatitlek, Chenega and Eyak leadership as well as former Mayor Tim Joyce of Cordova; GIS support from Samantha Greenwood.

LITERATURE CITED Bowyer, R. T., J. W. Testa and J. B. Faro. 1995. Habitat selection and home ranges of river otters in a marine environment: effects of the Exxon Valdez oil spill. Journal of Mammology 76:1-11. Chenega Corporation, 2008. Resource Management Plan for the Village of Chenega Bay [Draft]. Chenega Village Corporation, Anchorage, AK, 30 pp. Colt, S., Martin, S., Mieren, J., and M. Tomeo. 2002. Recreation and tourism in south- central Alaska: patterns and prospects. Gen. Tech. Rep. PNW-GTR-551. Portland, OR: U.S. Department of Agriculture, Forest Service, Research Station. 78 pp. ESRI (Environmental Systems Resource Institute). 2009. ArcMap 9.3. ESRI, Redlands, California. Exxon Valdez Oil Spill Trustee Council. 2010. Exxon Valdez oil spill restoration plan. 2010 update injured resources and services. Exxon Valdez Oil Spill Trustee Council, Anchorage, Alaska, USA. Fall, James A., Lee Stratton, Philippa Coiley, Louis Brown, Charles J. Utermohle, and Gretchen Jennings. 1996. Subsistence Harvests and Uses in Chenega Bay and Tatitlek in the Year Following the Exxon Valdez Oil Spill. Alaska Department of Fish and Game, Division of Subsistence Technical Paper No. 199. Fall, James A. and Charles Utermohle. 1999. Subsistence Harvest and Uses in Eight Communities Ten Years After the Exxon Valdez Oil Spill. Alaska Department of Fish and Game, Division of Subsistence Technical Paper No. 252. Juneau. Fay, G. Prince William Sound Tourism Economic Indicators. National Wildlife Federation Alaska Office. 2008. 31 pgs. http://www.nwf.org/News-and- Magazines/MediaCenter/Reports/Archive/2008/~/media/PDFs/Regional/Alaska/P RCA_PrinceWilliamSound_Tourism.ashx.

40 Gimblett. H. R. & R. M. Itami. 2006. Modeling the Distribution of Human Use in Prince William Sound Using the Recreation Behavior Simulator. Report for the Chugach National Forest. December, 2006. Hall, Troy E. and D. N. Cole. 2007. Changes in the motivations, perceptions, and behaviors of recreation users: Displacement and coping in wilderness. Res. Pap. RMRS-RP-63. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 37 p. Hennig S. and W. Menefee, 1995. Prince William Sound Recreation Project. Exxon Valdez Oil Spill Restoration Project Nos. 93065 and 94217. Lace, S., H.R. Gimblett & A. Poe. Applying an Agent-Based Modelling Approach to Simulating Spring Black Bear Hunting Activities in Prince William Sound, Alaska. In Gimblett, H.R. & H. Skov-Petersen (eds) Monitoring, Simulation and Management of Visitor Landscapes. University of Arizona Press. 2008. pgs. 295- 313. Murphy, K. A., L. H. Suring, and A. Iliff. 2004. Western Prince William Sound human use and wildlife disturbance model, Exxon Valdez Oil Spill Restoration Project Final Report (Restoration Project 99339), USDA Forest Service, Chugach National Forest, Anchorage, Alaska Poe, C. B. and S. Greenwood. 2010. Prince William Sound Human Use Hot Spots GIS Database and Spatial Analysis. Exxon Valdez Oil Spill Criminal Restitution Project Final Report. USDA Forest Service, Chugach National Forest, Anchorage, Alaska. Poe, A., H. R. Gimblett and R.M. Itami. 2010a. Evaluating the Recreation Service Recovery: Evaluation of Prince William Sound User Experience, Exxon Valdez Oil Spill Restoration Project Final Report. USDA Forest Service, Chugach National Forest, Anchorage, Alaska Scott, Cheryl, Louis A. Brown, Gretchen B. Jennings, and Charles J. Utermohle. 2001. Community Profile Database for Access 2000. Version 3.12. Alaska Department of Fish and Game, Division of Subsistence, Juneau, Alaska. State of Alaska, 2008. Housing Unit Method Manual: Populstion Estimate Instructions and Reporting Forms. Department of Commerce, Community, and Economic Development and Division of Community Advocacy. Juneau, Alaska, 23 pp. Stratton, Lee and Evelyn B. Chisum 1986 Resource Use Patterns in Western Prince Wrlliam Sound: Chenega in the 1960s and Chenega Bay 1984-86. Alaska Department of Fish and Game, Division of Subsistence Technical Paper No. 139. Juneau, Alaska. Tatitlek IRA Council, 2003. Tatitlek Village Natural Resource Management Plan. Indian Reorganization Act Council for Tatitlek Village. Tatitlek, AK. 26 pp. Twardock, P., and C.A. Monz. 2000. Recreational kayak visitor use, distribution, and economic value in Prince William Sound, Alaska USA. In Cole, D.N., McCool, S. F.,Borrie, W. T. and O’Loughlan, J. (comps.) Wilderness science in a time of change conference– Volume 4. Wilderness visitors, experiences, and visitor management.Proceedings RMRS-P-15-Vol-4. USDA Forest Service Rocky Mountain Research Station. Wolfe, P.E., B. Garber-Yonts & R. S Rosenberger, 2006. Monitoring and analysis of recreational boat use in sensitive wildlife areas in Prince William Sound, Alaska:

41 A simulation approach. ISSRM 2006 – 12th International Symposium on Society and Natural Resource Management. June 3-8, Vancouver, British Columbia Canada.

42 APPENDIX A. Specific questions asked during household interviews including directions given to observers. INTERVIEW APPROACH FOR SUBSISTENCE HARVEST PATTERNS IN PRINCE WILLIAM SOUND

Background The project will create a map database characterizing the distribution of subsistence harvest activities by season within Prince William Sound (PWS). This database will include data collected from harvesting households in the subsistence eligible PWS communities of Chenega Bay, Cordova, Tatitlek, and Whittier. Data will be collected through partnerships with groups or individuals within each of these communities. Harvest activities included in this database will include such resources as: bottom fish, deer, salmon, mountain goat, marine mammals, shellfish, plants etc. harvested commonly in the PWS region. The resulting database will be compared to the mapped distributions of both private and commercial recreation use activities throughout PWS to better understand the interaction between these activities. The results of this assessment will inform recreation management practices aimed at minimizing user conflict and displacement of subsistence users from traditional harvest areas in PWS.

Introduction These questions will be posed using classic household interview procedures (e.g., population census) in which the interviewer seeks an appointment with the heads of subsistence harvesting households within the target community. Answers to these questions will help describe the general locations important for harvest of resources by individual households, as well as the different seasonal patterns of harvest activity. They also will help assess the existing degree of possible conflict or competition between subsistence and other uses, as well as general observations about harvest conditions in PWS. Participants will be asked to identify harvest locations on a map of PWS with land areas based on State Fish and Game harvest units. These areas (UCUs or Uniform Coding Units) are used by Fish and Game to summarize harvest data for most species. The open water of PWS has been divided into sections based on State Fish and Game statistical polygons for Sportfishing guides. Hopefully, this map scale will allow respondents to feel comfortable participating without giving away their specific harvest locations. Both land and water map units each have a unique code which will be recorded by the interviewer on the associated table; any notes made on the map itself can be wiped clean at the end of each interview to alleviate concerns about specific locations being recorded.

In this document text in bold represents each of the questions that the interviewer will ask each respondent; the text in italics describes where that information is recorded on the attached table. The duration of interviews is anticipated to range from 20 to 40 minutes depending on the diversity of resources harvested by individual households.

The Draft Questions

43 Protocol: A set of 22 species flashcards will be provided to use as a visual aid and icebreaker for the interview. The interviewer should spread the cards out and ask the interviewee which species the household has harvested (question #1):

1) Which of the following resources do you usually attempt to harvest in a year? (The Species Flashcards will be used as a visual tool/icebreaker to record the species) A = Black Bear B = Brown Bear C = Crabs D = Deer E = Eggs F = Furbearers (river otter, mink, wolverine, etc) G = Mountain Goat H = Halibut I = Shellfish (clams, mussels, etc) J = Berries K = Harbor Seal L = Sea lion O = Sea Otter P = Plants (building materials, firewood, medicinal, etc.) R = Rockfish or Cod S = Salmon T = Other finfish (Trout, dolly varden, sharks etc.) U = Seabird (Murrelet, guillemot, terns, etc) V = Seaducks (Mergansers, Harlequins, Scoters, Goldeneyes) W = Waterfowl (other ducks and geese) X = Shorebirds Y = Shrimp Z = OTHER (please have them specify) [Each Recorded in Column 1: Resource- on the associated table]

Note: A table of these codes is provided for use in the interview

Following selection of harvested species, the interviewer should focus on one species at a time to map all harvest areas used for that species, and answer questions 2-6 before moving on to the next species harvested by the household:

2) For this resource, please review the map and note which areas (area numbers) your household uses? [Recorded in Column 2: Area #]

Note: Record the use for ONE species at a time to make sure all areas used for harvest are covered.

44 3) How many years total has your household been attempting to harvest this resource? [Recorded once in Column 3: Total Years of Harvest (Historically)] Note: This is the total historical use of harvest for this species. It may be concurrent or disjointed use, please but record the total number of years regardless.

4) Within the last five years, how many years has your household harvested the resource in this area? [Recorded in Column 4: Years of use]

Note: The subsequent questions (the gray shaded area) in this section refer to “current use” meaning the last five years of harvest of that species in that area.

5) For each area identified for this resource please complete the following 4 things:

A. The month or months your household uses this area for this resource? (If the answer is “June”, please ask if it’s the first part of the month or the last part) [Recorded in Column 5: Month harvested resource]

Note: June is singled out because we have a breakdown regarding season of harvest and June is the only month that is split between seasons.

B. The approximate total number of days your household uses this area to harvest that resource? [Recorded in Column 6: Total Days]

Note: If the interviewee harvested the species in two separate months (e.g. June and July), and harvested in June for 5 days, and harvested in July for 6 days, the total use is 11 days.

C. Has the household’s use (number of days) of this area increased or decreased? [Recorded in Column 7: Increase or Decreased ↑↓]

D. If your household’s use of this area has changed in recent years, what reasons best explain this change? A = General change in resource availability (seen more or less of that species in that area) B = Competing local/subsistence harvest (known individuals within your area) C = Non-local sport harvest (private individuals believed to be from outside of PWS)

45 D = Competing commercial harvest (large operations) E = Competing sport harvest (other individuals for private use, guides, or small operations) F = Presence of non-consumptive user (tours, individuals) G = Cost of Transportation H = Change in Regulations I = Other, please specify [Recorded in Column 8: Change in Use]

Note: We are relying on you the interviewer to code the response as best as possible into the above categories. The categories can be read as example reasons but we hope to not lead respondents to any categories. 6) Are there areas that you no longer harvest this resource and if so, please list those area numbers and the reasons that best explain why for each area listed? A = General change in resource availability B = Competing local/subsistence harvest C = Competing commercial harvest D = Competing sport harvest E = Presence of non-consumptive user F = Cost of Transportation G = Other, please specify [Recorded in Column 9: Areas No Longer Used]

7) Does your household share and/or receive this resource from other households in Prince William Sound?

After completing questions 2-6 listed above for each individual species/resource selected by the household, the interviewer should conclude with the following questions:

8) If you have seen changes in resource availability in PWS could you generally describe those changes?

9) Is there anything else regarding subsistence harvesting in PWS that you feel we should know about?

46